EP1470501A2 - Procedes et systemes de recherche et d'association de ressources d'information telles que des pages web - Google Patents
Procedes et systemes de recherche et d'association de ressources d'information telles que des pages webInfo
- Publication number
- EP1470501A2 EP1470501A2 EP20030709876 EP03709876A EP1470501A2 EP 1470501 A2 EP1470501 A2 EP 1470501A2 EP 20030709876 EP20030709876 EP 20030709876 EP 03709876 A EP03709876 A EP 03709876A EP 1470501 A2 EP1470501 A2 EP 1470501A2
- Authority
- EP
- European Patent Office
- Prior art keywords
- resources
- resource
- text
- relevant
- user
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 164
- 230000008569 process Effects 0.000 claims description 49
- 238000004364 calculation method Methods 0.000 claims description 20
- 230000011664 signaling Effects 0.000 claims description 14
- 238000007726 management method Methods 0.000 claims description 10
- 238000004458 analytical method Methods 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 8
- 230000009471 action Effects 0.000 claims description 7
- 230000008859 change Effects 0.000 claims description 6
- 239000012634 fragment Substances 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 5
- 238000006073 displacement reaction Methods 0.000 claims description 5
- 238000000354 decomposition reaction Methods 0.000 claims description 4
- 230000001419 dependent effect Effects 0.000 claims description 3
- 230000001960 triggered effect Effects 0.000 claims description 3
- 238000004821 distillation Methods 0.000 description 21
- 235000014510 cooky Nutrition 0.000 description 17
- 238000004422 calculation algorithm Methods 0.000 description 16
- 238000011144 upstream manufacturing Methods 0.000 description 15
- 238000003860 storage Methods 0.000 description 13
- 230000008901 benefit Effects 0.000 description 12
- 238000002360 preparation method Methods 0.000 description 11
- 230000000295 complement effect Effects 0.000 description 6
- 238000001914 filtration Methods 0.000 description 5
- 238000009826 distribution Methods 0.000 description 4
- 238000012423 maintenance Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 230000003042 antagnostic effect Effects 0.000 description 3
- 238000013459 approach Methods 0.000 description 3
- 238000011143 downstream manufacturing Methods 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- 230000002829 reductive effect Effects 0.000 description 3
- 230000004044 response Effects 0.000 description 3
- 230000002441 reversible effect Effects 0.000 description 3
- 241000209149 Zea Species 0.000 description 2
- 235000005824 Zea mays ssp. parviglumis Nutrition 0.000 description 2
- 235000002017 Zea mays subsp mays Nutrition 0.000 description 2
- 230000002457 bidirectional effect Effects 0.000 description 2
- 235000005822 corn Nutrition 0.000 description 2
- 230000009193 crawling Effects 0.000 description 2
- 230000007423 decrease Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000000670 limiting effect Effects 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000036961 partial effect Effects 0.000 description 2
- 101100264195 Caenorhabditis elegans app-1 gene Proteins 0.000 description 1
- 241000854350 Enicospilus group Species 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000008014 freezing Effects 0.000 description 1
- 238000007710 freezing Methods 0.000 description 1
- 238000011068 loading method Methods 0.000 description 1
- 239000010813 municipal solid waste Substances 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000013138 pruning Methods 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 230000036962 time dependent Effects 0.000 description 1
- 238000011282 treatment Methods 0.000 description 1
- 230000003313 weakening effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/955—Retrieval from the web using information identifiers, e.g. uniform resource locators [URL]
- G06F16/9558—Details of hyperlinks; Management of linked annotations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/957—Browsing optimisation, e.g. caching or content distillation
- G06F16/9574—Browsing optimisation, e.g. caching or content distillation of access to content, e.g. by caching
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y10—TECHNICAL SUBJECTS COVERED BY FORMER USPC
- Y10S—TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y10S707/00—Data processing: database and file management or data structures
- Y10S707/99951—File or database maintenance
- Y10S707/99952—Coherency, e.g. same view to multiple users
- Y10S707/99955—Archiving or backup
Definitions
- the present invention relates generally to methods and systems for managing resources such as Web pages accessible by the Internet, or any other type of document, aiming on the one hand to improve the obtaining of “close” resources. of given resources, in terms in particular of centers of interest for the user, and aiming on the other hand to allow the user, in a particularly simple and intuitive manner, to make himself associations between resources, especially to take advantage of it when obtaining close resources.
- An object of the present invention is to propose methods and computer systems for searching for resources (in particular web pages, various computer documents) "close” to given resources (this notion of proximity being explained below), as well as management methods. associative of resources.
- the invention aims to characterize pieces of information in relation to new pages that appear on the Web, thus opening the way for multiple new applications for dynamic content management in relation to the browsing context of the user. . More specifically, the invention aims to have each piece of information associated with links to relevant web pages which characterize it and which are automatically kept up to date. We can thus characterize non-textual information, such as photos, sounds and animations (in Flash, etc.) and dynamically select the elements to present to the user according to the context of his navigation which is also characterized by sets relevant web pages.
- non-textual information such as photos, sounds and animations (in Flash, etc.
- This approach is particularly, but not exclusively, suitable for magazines in the art of living, fashion and in all the other "taste" areas where it is difficult to characterize by keywords the interest that a subscriber takes in information (when for example it represents music, an art object, a culinary dish, etc.).
- Another object of the invention is to associate elements of information with other targeted elements, such as targeted advertisements, in exchange for an innovative service of associative memory offered to Internet users.
- the invention also aims to take advantage of modern user interfaces to create, in a particularly simple and intuitive manner, associations between information resources (web pages, or document files) in particular within the framework of the above objectives. above.
- the invention proposes, according to a first aspect, a method for determining additional resources relevant to a given set of starting resources, characterized in that it comprises the following steps a) identifying a set of citing resources constituted by all the resources having a link to at least one of the starting resources, b) forming a set of candidate resources constituted by the set of resources cited by the citing resources, c) for each candidate resource, calculating a score of relevance of candidate resource between said candidate resource and the set of starting resources on the basis of the existence of links located in the citing resources and directed towards the candidate resource and towards the starting resources, and also on the basis of resource relevance scores citing assigned to each citing resource, d) for each citing resource, recalculate a relevance score citation of a citing resource on the basis of the existence, in the citing resource in question, of links to the candidate resources and also on the basis of the candidate resource relevance scores assigned to the candidate resources in step c), e) if necessary repeat step c) and if necessary one or more times
- the relevance score calculation performed in step c) advantageously comprises the calculation of a plurality of sums of relevance scores of citing resources, each sum comprising only the relevance scores of the citing resources comprising a link to a given resource constituted by the candidate resource or a starting resource.
- the above method also comprises the calculation of at least one sum of relevance scores of citing resources, each sum comprising only the relevance scores of the citing resources comprising a link to one of a set of at least two given resources, this set comprising the candidate resource and at least one starting resource.
- the invention proposes a method for determining additional resources relevant to a given set of starting resources, characterized in that it comprises the following steps a) identifying a set of cited resources constituted by all the resources having a link from at least one of the starting resources, b) forming a set of candidate resources constituted by the set of resources citing the cited resources, c) for each candidate resource, calculating a candidate resource relevance score between said candidate resource and the set of starting resources on the basis of the links located in the candidate resource and in the starting resources and directed to the cited resources, and also on the basis of resource relevance scores cited assigned to each of the cited resources, d) for each cited resource, recalculate a relevance score d e resource cited on the basis of the existence, in the cited resource in question, of links from the candidate resources and also on the basis of the candidate resource relevance scores attributed to the candidate resources in step c), e) repeat where appropriate step c) and if necessary one or more times step d) then step
- the invention further provides a system for browsing among information resources, each resource comprising at least one link that can be activated in a first mode by an input device to cause access to another information resource designated by a resource identifier associated with this link, characterized in that at least certain resources include at least one link that can be activated in a second mode using an input device to send new search resources to a search engine information a search query containing the resource identifier associated with the link in question.
- This system has the following preferred but optional aspects: * the input device is able to activate the link simultaneously in the first and second modes. * the activation of the link in the second mode is capable of causing the display of a pre-existing request, to which the resource identifier associated with the link in question is likely to be added.
- the activation of the link in the second mode is capable of displaying, in addition to the pre-existing request, the information resource designated by said resource identifier.
- the invention also provides a system for searching for new information resources from existing information resources, characterized in that it comprises a search engine based on the analysis of links between different resources and accepting as input request comprising a series of resource identifiers, an identifier selection means capable of storing a set of resource identifiers (URI) selected one after the other by a user, and a request generator means that can be activated by the user to generate a request containing all the identifiers previously selected for the search engine.
- URI resource identifiers
- the selection means is able to store the identifiers selected in a non-volatile manner, so that the selection means can be implemented spaced over time with a view to generating the same request.
- the invention further provides a method of searching for new information resources from existing information resources, characterized in that it includes the implementation of a search engine based on the analysis of links between different resources and accepting as input a request comprising a series of resource identifiers and in that it comprises the following steps:
- URIs resource identifiers
- a method of searching for new information resources from existing information resources is also proposed, characterized in that it includes the implementation of a search engine based on the analysis of links between different resources. and accepting as input a request comprising a series of resource identifiers and in that it comprises the following steps:
- each group of resource identifiers is represented by a graphic object on a display device of the user, and in that said signaling is carried out at least by change of appearance of this graphic object.
- the invention further provides a method of managing resources in a computer system provided with a display screen and an input device for moving and actuating a cursor such as a mouse, each resource having a displayed representation. on the screen of so that it can be moved using the input device, a method characterized in that it comprises the following steps:
- the method further comprises, following the identification of a given resource in a resource consultation process, the following steps:
- resources include files.
- the resources include resources accessible by a network such as the Internet.
- the identification of a given resource is obtained by a process of identification of similar or relevant resources with respect to at least one starting resource.
- the signaling step comprises the ordered signaling of at least part of said several associated resources.
- the associative resource management memory is contained in a server accessible from a plurality of individual stations in which the displacement step can be implemented.
- the invention also proposes a method for identifying from a text resource, a part of said resource capable of constituting a significant request for a search engine, characterized in that it comprises the following steps:
- the step of implementing the resource distillation method is carried out only with portions of text selected as preponderant, where the quoting portions of text are the portions of text which include at least one word in common with the one or more predominant parts of text, where a link is created from each part of citing text to the part or parts of predominant text, where the parts of text containing at least one word also contained in the parts of citing text are identified, to form a group of co-cited text parts, and where is temporarily created a link from each citing text part to each co-cited text part with which said citing text part has at least one word in common.
- Parts of text are typically sentences.
- the invention proposes a method for managing information resources such as web pages in a computer system comprising a user station provided with a display screen, each resource having an identifier (URI) allowing its access from the user station, process characterized in that it comprises the following steps:
- step (b) includes the selection of other resources most relevant to the implementation of step (c).
- step (a) is implemented for a plurality of second resources belonging to a group, and in that step (b) comprises the identification of other resources relevant to all of the second resources of the group.
- step (b) is triggered by carrying out step (a).
- step (b) is implemented after the access provided in step (c) to determine if the other resource to which it has been accessed is another resource relevant to the second resource.
- step (b) is implemented by supplying an identifier of the second resource to a server for determining relevant resources.
- step (b) is implemented by identifying other relevant resources with respect to at least one intermediate resource (spot) with respect to which the second resource is predetermined as being relevant.
- the method further comprises the display, in the vicinity of a resource display area, of representations of links to at least some of the first resources, the intermediate resources, and of resources relevant to the intermediate resources.
- step (a) is implemented by action using an input device on graphic objects representative of the first and second resources.
- the invention also proposes a method for identifying information resources accessible by recent links (such as web pages), relevant to at least one given resource, characterized in that it comprises the following steps:
- the invention proposes a method for allowing access by a user to relevant information entities from a starting information entity, each information entity being accessible by a identifier (URI), characterized in that it comprises the following steps: a) providing at least one similar information entity, with content similar to that of the starting entity, and determining the identifier of the or each similar information entity, and b) determining from the or each similar information entity identifier a set of one or more information entity identifiers relevant to the or each information entity similar.
- URI identifier
- the method further comprises the following step: c) allowing the user access to at least certain relevant information from their respective identifiers.
- the method further comprises the following step: d) from the identifiers of relevant information entities and from a given set of additional information entities, selecting the additional entities most similar to the information entities relevant.
- the method includes an additional step of sorting the relevant information entities by degree of relevance.
- the sorting step is preceded by a step of calculating a relevance score with respect to the or each similar information entity for each of the relevant information entities.
- each information entity is made up of a page fragment written in standardized markup language, or of such a page as a whole.
- each identifier is made up of a uniform resource identifier (URI) for the fragment or page.
- URI uniform resource identifier
- step a) is carried out by selection by the user of one or more information entities similar to the starting information entity.
- step a) is carried out by implementing a process for automatically determining similar information entities.
- step a) is carried out by implementing a process of automatic determination of similar information entities, followed by a selection by the user of one or more similar information entities from among the entities d 'similar information determined by said process.
- step b) is carried out by implementing a process of automatic determination of relevant information entities.
- the process of automatic determination of relevant information entities includes the analysis of an identifier graph structure constituted by the identifiers of information entities and by the identifiers designated by links that can be activated by the user contained in said information entities.
- a method for determining relevance scores of text units such as sentences in a text document comprises the following steps:
- the invention also provides a method for determining relevance scores of text units such as sentences in a text document, characterized in that it comprises the following steps: - decomposition of the document into a plurality of text units ,
- Method for determining scores assigned to words or groups of words contained in text units such as sentences in a textual document characterized in that it comprises a step which consists in adding the relevance scores, determined by one of the above methods, text units in which said words are found.
- FIGS 1 to 7 of the accompanying drawings illustrate different steps implemented in the present invention.
- Resource Information resource such as a Web page, part of a Web page, a document, or even an XML element. Each resource can itself be made up of resources, thus forming a tree structure. Current resource: Resource accessed by the user at the current time when browsing (this is in particular the web page viewed in the main browser window).
- URI Uniform Resource Identifier
- URL universal resource locator
- Link URI placed in a resource. In general, by clicking on a link, the user can access the resource pointed to by him.
- Quote (A first resource çjte a second resource): The first resource has a link to the second resource.
- Popular Said of a resource which is accessed by a large number of users (for example on the Web) from its URI.
- Private resource Resource which is not accessible by a large number of users (in particular which is not published on the Web or which is not widely known).
- Associative storage Adding a link to a first resource, on a second resource, in order to be able to find the first resource using the associative search process - Associative search: In order to find a first resource, access a relevant resource in relation to a second resource to which a link to the first resource has been added. Added link: URI inserted by the user in a set of associated links.
- Suggested spot Spot presented by the system as a priority because it contains the most relevant associated links in relation to the current context.
- a spot is made up of:
- Associated links are made up of given associated links and completed associated links,
- one or more set of links (in particular links added by the creator of the spot and links added by users of the spot), offered to the user as part of the process of associative research,
- Area of relevance of a spot All the resources designated by the associated links of this spot.
- Associated links given Associated links explicitly specified (by the one who creates or publishes the resource with which said set is associated, or by the one who creates a spot for this resource).
- Associated links determined automatically (in particular by means of a relative distillation algorithm described in the present description).
- Associated link score Relevance score compared to all the associated links given. This score can be calculated by a relative distillation algorithm such as one of those described in the present description.
- Authority score Score of relevance of a resource in relation to a set of given associated links.
- Pivot score Relevance score of a resource citing other resources, representing the relevance of the cited resources in relation to a set of given associated links.
- Non-contextual score Relevance score does not depend on the context.
- Contextual score Relevance score depending on the context.
- Non-contextual spot in relation to a resource (or a set of resources) in question Spot whose associated links include the URI of the resource in question (or at least some of the URIs of the resources in question) with a score ( or an average of scores) greater than a given threshold or which is selected so as to maximize it (or the (see the spot selection method described in this description).
- Contextual spot Spot whose associated links are the most relevant to the context.
- Spotserver Server on the Internet providing the association between associated link and spot.
- Relevant region of a resource Part of a resource containing at least one relevant link and not containing an irrelevant link.
- first page page memorized by the user so that he can find it easily
- second page page used by the user as a storage medium (to memorize an association with the first page, which will be referred to below as "to memorize the first page” for the sake of brevity)
- current page page presented at the current time in the main window of the Internet browser.
- the first page can be a private resource such as a document (text, multimedia or other document) which belongs to it].
- the system allows the user to add a link to a first page on any second page (or in the vicinity of the second page; we will use the expression "on the second page” below for the sake of brevity ).
- the user accesses the pages using a browser with the system-specific extension (or via an intermediate Web server).
- Adding a link can be done for example by drag and drop: the user grabs a handle representing the first page and places it on the second page; for example the added link is then presented by the system as a thumbnail in the style of a "post-it" in the place where it was deposited, or in a window adjacent to the main browser window (or in a frame adjacent to the frame showing the original web page). He can also drop it on an icon representing the second page (for example
- the step consisting in thus adding a link, on a second resource, to a first resource (in order to be able to find it by the method described in this report) is called associative storage. in his favorite links).
- the system then stores the relationship with the user in question, the association between the link on the first page and the second page in question.
- the user only has to access any page (accessible by the system) which is relevant to the first page : this is the associative research process. Note that during the associative memorization step, the user can increase his chances by adding a link to the first page on several second pages.
- the links added are implicitly bidirectional.
- the system can assimilate it to the second page (s) on which (if any) the user had added a link to this private resource, and present the other first pages that he had also added on this (these) second (s) page (s).
- the associative memorization step can be automated (or be computer assisted). Indeed, adding a link to a first page on a second page can be (semi-) automated according to the following steps: I - determining keywords or main sentences of the first page (contained in the page or associated with it - for example delimited by “meta-tag” tags),
- III - take at least a subset (for example the N best ones according to the search engine) to use them as second pages,
- IV - add a link to the first page on these second pages.
- step I various techniques for automatically extracting keywords or main sentences from a text already exist.
- Any such page is already or should be taken into account by the system.
- the user will prefer to choose a popular page to speed up the search.
- the system is equipped with a crawler whose purpose is precisely to take into account as much as possible accessible pages (especially on the Internet) and which interest the user.
- the relative distillation method (described below) can be used to determine the relevance score of the sentences with respect to said preponderant sentences.
- Text units comprising at least one word in common with the predominant unit (or all of the units) are identified, to form a group of quoting text units.
- a link is created (temporarily) from each citing text unit to the preponderant text unit (or all of the units).
- Text units containing at least one word also found in the quoting text units are identified to form a group of co-cited text units.
- a link is created (temporarily) from each citing unit to each co-cited unit with which said citing unit has at least one word in common.
- the set of identifiers of the relevant text units constitutes the URIs of the request. 5
- the set of identifiers of the citing text units constitutes the set R ⁇ .
- the set of identifiers for the co-cited text units constitutes the set R ⁇ + , and so on.
- Step a determine the relevance score of second candidate pages in relation to the current page
- Step b select the (or a number of) second pages having (if applicable) a sufficient relevance score
- Step c present to the user the (URI of) first pages of the links that he had added on the second pages which were selected in step b; optionally also present to it the (URIs of) second pages themselves.
- the user can overlay or insert an annotation (or any resource such as an icon or other graphic object), which then plays the role of first page within the meaning of the present process.
- the system presents the second page or pages which have been selected, also presenting their annotations (or the resource which has been added to them).
- R consists of pages 9 of the request.
- R ⁇ is the set of pages that contain a link to a page at least 10 of the request.
- R '+ is the set of pages pointed (cited) by at least one page of R ' .
- R '+ " is the set of pages that cite at least one page of R ' + (R ⁇ z R ⁇ + ⁇ ).
- R + is the set of pages cited by at least one page of the query (R) .
- R + ⁇ is the set of pages which cite at least one page of R + .
- R + '+ is the set of pages cited by at least one page of R + ⁇ (R * a R + ⁇ ).
- the system implements a “relative distillation” process comprising at least one of steps a and a 'following.
- This step is made up of step a and / or step a '(see below 7)
- the system has in memory the relationship between user, second page (on which the user in question has added links) and first page (link added by the user in question on the second page in question).
- the system can first of all determine the set of second candidate pages for the current user to perform step a, then in step c find the added links to present to the user.
- link added on a second page we mean that we include this scenario where there is a resource added to the second page.
- page means “page URI”.
- Step a-1 Identify the set R " of pages which have at least one link to R;
- Step a-2 Find in memory all the second candidate pages for the current user and perform the intersection between the set R '+ of the pages pointed by the pages of R " (note that R is in the set R "+ ) and all of the second candidate pages for the current user;
- Step a-3 For each page of the set resulting from step a-2, calculate its relevance score (authority score) compared to R. (Note that this step includes the identification of all the pages of R '' with at least one link pointing to at least one subset of the set resulting from step a-2 - see the section "Selection of spots"). Step a ':
- Step a'-l Identify the set R + of the pages pointed by R;
- Step a'-2 Find in memory all of the second candidate pages for the current user and perform the intersection between the set R + ⁇ of the pages having at least one link to a page of R + (note that R is in the set R + ⁇ ) and the set of second candidate pages for the current user;
- Step a'-3 For each page of the set resulting from step a'-2, calculate its relevance score (pivot score) compared to R. (Note that this step includes the identification of the set pages of R + '+ pointed to by at least a subset of the set resulting from step a' -2).
- the calculation of the relevance scores in steps a-3 and a'-3 can be carried out using one of the equations presented below in the “Selection of spots” section, which also describes improvements to the process presented here. -above.
- the scores are refined by successive iterations. During these iterations, the pivot pages in step a and the authority pages in step a 'also acquire relevance scores (pivot scores and authority scores respectively).
- step b In addition to the second candidate pages (that is to say in addition to the URIs of the pages of R '+ in step a and / or of R + ⁇ in step a') determined as described above, we can then also include, in the resulting set provided in step b, the pivot pages of step a and the authority pages of step a '(since they now have relevance scores). Furthermore, the weights of links between close pages 13 are lessened to improve the results further. The system can therefore select the second most (or sufficiently) relevant pages in step b and perform step c to present the user with their added links.
- the results obtained by the relative distillation process can be memorized (and then maintained - see the “Spot maintenance” section below) in order to avoid recalculating them when accessing the current pages already processed.
- the system maintains, in a second memory, the scores of the second pages relative to the current pages in the cases where these scores are greater than a given threshold. For a current page already processed, the system response is then almost immediate.
- step a is modified as follows:
- Step a consult the second memory to know if the second pages most relevant to the current page have already been memorized (and if these data in memory are sufficiently fresh), if necessary go to step c, if not determine and memorize the relevance score of second candidate pages with respect to the current page.
- the system memorizes (then maintains - see the “Spot maintenance” section below) the necessary data without waiting for a user to access a current page; storage is triggered by the use by the user of a new second page (as an associative storage medium).
- the system starts from each second page to construct R ' and R ' + (and R ⁇ + ⁇ ) and / or R + and R + ⁇ (and R + '+ ), calculates by relative distillation the relevance scores of all the potential current pages, and stores them in a second memory (it is a reverse memory capable of providing, for each potential current page, the second relevant pages).
- the system maintains a first memory containing the links added by user and second page.
- the system selects in the second memory the second pages - from the second pages used by this user as storage medium 16 - which have the highest relevance scores with respect to said current page, then find (in the first memory) the links added by this user on these second pages.
- the method comprises the following steps 17 .
- Step ml Perform at least one of the steps ml-1 and ml-1 ', then carry out the step ml-2:
- Step ml-1 - identify the set R ' of pages which have at least one link to R;
- the data in the second memory is not per user and can thus be used by all users.
- Steps ml and m2 describe the associative memorization process, steps a, b and c describe the associative search process.
- the ml step is performed only for new second pages, while the m2 step is performed each time a second page used by a user, whether new to the system or not.
- - identify the set R + ⁇ of potential current pages pointing to at least one page of R + ;
- Step m 1-2 store, in a second memory, the URIs of the pages having a sufficient relevance score with respect to R, in relation to R, so that from the URI of each of said pages having a sufficient relevance score with respect to R we can find 19 (the second page) R as well as said sufficient relevance score;
- Step m2 (in parallel with step ml) memorize in a first memory, for each user and each second page, the added links that said user has added to said second page;
- Step a is no longer necessary since the scores are already in memory.
- Step bm Select in the second memory a number of second pages 20 , from among the second pages used by said user (indicated in the first memory), for which the relevance scores of said current page are the highest (if they exist);
- Step c (unchanged): Find in the first memory the links added by said user on the second pages selected in step b-m and present them to said user (with optionally the second pages on which they were added and sorted).
- the pivot pages in the ml-1 step and the authority pages in the ml-1 'step also acquire relevance scores (respectively pivot scores and authority scores) and can thus be included in the resulting set supplied in step ml-2 (in addition to the URIs of the pages of R '+ in step ml-1 and / or of R + ⁇ in step ml-1').
- the weight of links between close pages is reduced to improve the results (see the "Filtering" section).
- the URIs of the second pages relevant to a potential current page are already sorted by relevance score.
- the group or groups of second pages that he had already formed, if necessary for the first page, are offered to him by the system and he can then choose one or more of these groups in which to insert said new second page, or otherwise he can create a new group formed by the only new second page.
- each group is treated by the system as a relative distillation request. Similar to the last method described 21 , for each request R (that is to say for each group of second pages) the system identifies and stores (then maintains - see below the "Spot maintenance" section) the pages potential currents that have a sufficient relevance score, and thus forms a reverse memory capable of providing, for each potential current page, the most relevant queries (that is to say the most relevant groups).
- associative memorization includes the following steps:
- Step ml Perform at least one of steps ml-1 and ml-1 ', then carry out step m 1-2:
- Step ml-1 ' - identify the R + set of pages to which at least one page of R has a link;
- Step m 1-2 memorize, in a second memory, the URIs of the pages having a sufficient relevance score with respect to R, in relation to R, so that from the URI of
- R represents a request formed by one or more resources whereas before R represented a single resource (a single second page). each of said pages having a sufficient relevance score with respect to R, it is possible to find 22 R as well as said sufficient relevance score;
- Step m2 (in parallel with step ml) memorize in a first memory, for each user and request, the links added (to first pages);
- Step m2 memorize in a first memory, for each user and request, the links added (to first pages);
- Step bm Select in the second memory a certain number of requests, among the requests (groups) used by said user as associative storage medium (indicated in the first memory), for which the relevance scores of said current page are the most high (if they exist);
- Step c Find in the first memory the links added by said user on the requests selected in step b-m and present them to said user, with optionally:
- step ml-2 the pivot pages in the ml-1 step and the authority pages in the ml-1 'step also acquire relevance scores (respectively pivot scores and authority scores) and can thus be included in the resulting set provided in step ml-2 (in addition to the URIs of R "+ pages in step ml-1 and / or R +" in step ml-1 ').
- the weights of links between close pages are also reduced here to improve the results (see the "Filtering" section).
- step bm the system provides a set of selected requests. It would be advantageous to refine the selection so as to present to the user (s) the requests 24 which are the most relevant in relation to the browsing context of the user. This is what we will now describe.
- the browsing history of a user is modeled using a “context stack”, where each link (which can be presented to the user) is associated with a relevance score for each browsing level, and when a link is non-existent, we assimilate it to a link whose score is zero.
- the system adds a level to the context stack.
- the system unstacks a level.
- the context score is an average of the non-context scores 25 at each level of the context stack, weighted as a function of depth.
- the context score at a certain level is the weighted average of the non-context score at this level and the context score at the previous level.
- the contextual score at the last level is: lambda.r + (1 - lambda).
- the system selects those which are closest to the context, i.e. those for which the scores of the URIs stored at the 'step ml-2 are the closest to the context scores for the user in question.
- the system calculates the sum of the products, for each URI of the request, of the score (non-contextual) of the request with the context score for the user in question.
- Step b-m is thus replaced by the following step b'-m:
- Step b'-m Select in the second memory a certain number of requests, among the requests (groups) used by said user as associative storage medium (indicated in the first memory), for which the relevance scores of said current page are the highest (if they exist) and for which the relevance scores of the potential current pages are the closest to the contextual relevance scores.
- cookie system allows site servers of an Internet domain (ie domain name or IP address) to recognize a user (i.e. his computer) when he accesses web pages belonging to a same Internet domain.
- the process described here allows a server, which implements our process - we will call it client server (CLI) - to recognize even users who navigate from one site to another who are not part from the same Internet domain, even moreover if in their navigation these users pass through sites which do not implement our process.
- CLI client server
- Each web page of a client server site contains a frame whose address is that of a centralized server (URS) which manages our user recognition process
- the centralized server and each client server each have a cookie stored in the user's computer (note that the time of creation of these cookies can be used to estimate the reliability of user recognition); 3- The client server communicates with the centralized server directly.
- New user for the client server and for the centralized server is a new user for the client server and for the centralized server:
- CLI sends USER the content of the http: //CLI.com / ... framework as well as a new cookie containing ID- 'ABCDEF "as a replacement from the previous User already known to the centralized server and the client server:
- URS responds 6. (Optionally, CLI can request and / or receive from URS additional data for ID- 'ABCDEF ")
- CLI sends the content of the http: //CLI.com / ... framework to USER (if applicable after receiving the data in step 6.)
- the context score is an average of the non-context scores at each level of the context stack, weighted as a function of depth.
- r being the non-contextual score at the last level and ⁇ the contextual score at the previous level, its value after following a link is: r t l- » ⁇ r, + ⁇ r t 21
- the links presented to the user are those who have the highest context score.
- the context stack can be displayed in the URS frame (the first frame) introduced above. So the user can see which pages were used in the calculation of the pages to be displayed. He can click elements of the stack to go up levels, and an "Erase" button allows to empty the context stack.
- the context stack is stored, for each user, in the centralized server (URS), with the user identifier.
- URS centralized server
- the server of the client site can then display in the page the links which have the best score.
- the user sends an http request to open a page
- the pages to be sent can be selected by taking only those which have a score above a certain threshold, for example half the threshold required for a page to be displayed to the user.
- the client server transmits to the centralized server (URS) the non-contextual scores of the page in question and the user identifier
- the client server selects the links with the highest score and presents them to the user.
- the current context 30 must contain context information for each part of the page displayed, so when the page sends its non-contextual scores, it sends as many as there are parts, and URS responds with a context for each part. To avoid certain problems (see following points) you also need a default context, which represents the page itself and its parts and which accumulates all the scores of all the links
- URS looks at the user's context for part 1, checks that the source page (http://CLI.com/main.html) corresponds to the last level of the context stack for this user (otherwise he would have ignored the part number and taken the default part "D"). Then it calculates, for each part of the new page the new contextual scores.
- URS having received the message (6), can respond to the user's message (4) (presenting him the new context stack and the ⁇ ERASE> button).
- URS also responds to CLI message (6) by sending it the user's real identifier (678910), as well as the contextual scores.
- the first pages can in particular be advertisements which (thanks to the capacities of the system which one has so far) are automatically selected compared to the context.
- spots The groups created by the supplier user and proposed by the system to the end user are called "spots”.
- the provider user manipulates and uses the spots as described so far for groups of second pages.
- the end user can use a spot as a storage medium by making a personal version and adding a link to a first page (this is described below).
- the main advantage of this approach is to give the possibility of creating new spots (and the costly score calculations they involve) to certain users only (these are supplier users) and to offer the memory / search function. associative via pre-existing spots (which is not expensive in terms of machine resources) to all users.
- the spot is offered in a window adjacent to the main browser window, like existing systems providing "related links", however unlike these existing systems
- the system of the invention has relevant links determined according to a relative distillation process (detailed below),
- the navigation context taken into account by our system is not necessarily only the current page, but can include all of the resources recently accessed by the user (using the system) and which are relevant to the resource current 31
- the spots serve as associative memory for supplier users; in fact, when a spot is presented to an end user, the links to the first pages (or other added resources 3 , as described above) added by the supplier user who created the spot are presented to said end user 33 ,
- the spots serve as associative memory for the end users; indeed, when the end user adds a link to a first page on a second page (as described so far), in reality he adds a link to his personal version of the spot proposed for this second page or for the current context.
- the provider user creates a new resource or chooses an existing resource (for example a web page he has just accessed, or a particular element contained in a page %) to make it the reference resource for a new one spot.
- This spot has the advantage not only of using it for himself but also of seeing it offered to end users.
- a link on the reference resource (encouraging the user to click) is included in the presentation of the spot, the reference resource is thus promoted to end users.
- its added links (such as advertisements) on this spot will be shown to end users.
- Second usage scenario On the Web, the end user "falls" on a first page (or other type of resource) so interesting that he would like to memorize it so that he can find it easily and fall back on it spontaneously when he accesses resources relevant to it.
- the user visits a (at least one) second page, which is relevant to the first, and for which he knows that a spot is offered, or otherwise he chooses a web page which is popular because it is thus more likely that a spot be proposed for her, and on the spot that is proposed for this second page he adds a link to this first page (for example by selecting a graphic object representing the first page and by dragging and dropping it on the second page, as described at the beginning of the description).
- this added link will then be spontaneously presented to him each time this same spot, or a nearby spot, will be offered to him for the current context of his navigation.
- Third usage scenario The end user wants to memorize a private resource (such as a document which belongs to him and which is not published on the Web).
- the private resource plays the role of front page here.
- He accesses a (second) page that is relevant to his private resource (and which is preferably popular, or for which he knows that a spot is offered) and he adds a link to his private resource (c ' that is, he inserts this link in his personal version of the spot proposed for this second page).
- This spot will not necessarily be offered, but rather, among all the spots whose associated links point to resources forming the current context, the spot in which these associated links have the highest relevance scores (or spots where these associated links have the highest relevance scores).
- the selection of the spot (s) is described in the section "Selecting a spot”.
- the system Before the actual spot (s) selection step, the system must obtain the set of “completed associated links” from the set of “given associated links” (which are given by the supplier user, such as described in the first usage scenario). Complete the associated links:
- the set of resources pointed to by the associated links given is the R request.
- Step 1 Identify the set R " of resources that have at least one link pointing to an element of R.
- Step 2 Identify the set R " of resources pointed to by the elements of R " (note that R "+ includes R).
- Step 3 For each resource of R " calculate its authority score in relation to R. (This step can include the identification of part of the resources of R “ + " having a link pointing to a resource of R _ + ) 38 .
- the calculation of the scores in step 3 can be done by calculating, for each resource of R "+ , the ratio between
- Step 4 The pivot score of each element of R " is obtained by taking the sum of the authority scores (calculated in step 3) of the elements of R " + to which it points. Pivot scores are normalized (so that their sum becomes 1).
- Step 3 takes into account the pivot scores so as not to consider all the elements of R " on an equal footing (the resources of R " pointing to resources having a higher authority score will thus have a greater influence).
- the cardinalities used to calculate the authority scores are thus replaced by weighted cardinalities. That is, each pivot resource, instead of counting for one, counts proportionally to its pivot score. (The equations are detailed below).
- Step 3 then includes taking into account the resources of R “+” pointing to the resources of R “+ having the highest authority scores, in addition to R " (a method optimizing the taking into account of R "" is described later).
- step 3 we can optionally perform step 4 again, and so on until convergence, that is to say until the difference between the results obtained in the last iteration and those obtained in the previous iteration is negligible (in general, less than 10 iterations are enough).
- Variant for step 2 To form R "+ , instead of taking all the links contained in the resources R " the system will only take the links located in the relevant regions of the resources of R " . As these relevant regions do not can be determined that from the moment the pivot scores of the links they contain are known, this variant will only be implemented from the first iteration, i.e. after having performed step 4 the system will iterate by resuming from step 2 rather than from step 3.
- Each link owned by a resource of R " (or R " + " ) is associated with a weight equal to the complement of the proximity of the two resources linked by this link.
- R " or R " + "
- the system will calculate the authority scores using no longer the sum of the pivot scores, but the sum of the pivot scores multiplied by their weights (this is detailed and illustrated by an example below).
- the proximity of the two resources linked by the link in question is obtained by calculating the ratio between
- pivot pages in the ml-1 step and the authority pages in the ml-1 'step acquire also relevance scores (respectively pivot scores and authority scores) and can thus be included in the resulting set provided in step ml-2 (in addition to the URIs of the R "+ and / or R +" pages) .
- the system forms a reverse memory capable of supplying, for each potential current resource corresponding to an associated link, the most relevant requests (i.e. spots most relevant).
- associative memorization now includes the following steps:
- the mO step is performed independently of the other steps.
- the ml step is performed only for requests, not already known by the system or not sufficiently fresh, submitted by a supplier user, while the m2 step is performed for each use of a request (i.e. a spot) as an associative storage medium by a supplier user or an end user.)
- Stage mO memorize (in a third memory) the rights to use spots for each user.
- Step ml corresponds to completing the associated links as described above.
- Step ml-2 store, in a second memory, the URIs of the resources having a sufficient relevance score with respect to R, in relation to R, so that from the URI of each of said resources having a sufficient relevance score with respect to R, 40 R can be found as well as said sufficient relevance score;
- Step m2 (in parallel with step ml) memorize in a first memory, for each user and request, the added links (to first resources);
- Step bm Select in the second memory a certain number of requests, among the requests (spots) that said user has the right to use (indicated in the first memory), for which the relevance scores of said current resource are the most high (if they exist) and for which the relevance scores of the associated links are closest to the contextual relevance scores for said user;
- Step c Find in the first memory the links added by said user on the requests selected in step bm, as well as the links added by their creators (if they are different from said user), and present them to said user, with optionally :
- RJ is the set of URIs for pages with a link to page R
- FfRJ is the set of URIs for pages to which R has a link.
- pi be the probability 42 that a random (web page) author puts a link on P / on a page.
- p2 be the probability that a random author puts a link on P 2 on a page.
- PJ is the set of URIs for pages with a link to page P ,.
- F (P) is the set of URIs of the pages to which P has a link.
- the relevance of a page to a set of pages can be defined by the “amount of common reasons” for being interested in all of these pages.
- Pj and P_, are web pages, although the methods described are much more general, as already mentioned.
- the system can be based on an analysis of traces of copy-paste (or cut-and-paste) of fragments of information made by users (in the context of the creation and manipulation of information resources), to automatically suggest other fragments that are likely to enrich these resources.
- traces can in fact be likened to links. For example, when copying part of a web page into a document, the system is able to deduce and remember the existence in the document of a link to the web page, and the same mechanisms described here can then be applied.
- the method described here can advantageously be applied by assimilating the links of a resource to another resource to links of a user to a resource he likes (that is to say a resource that interests him). We can thus determine the quantity of common reasons (between several resources) to be liked by users. This can notably be used to categorize these resources.
- the probability of being interested in one (or some) page (s) is approximated by counting the number of pages that have a link on it (s) and dividing this number by an estimate of the number of pages that could have been to have.
- the algorithm allows, having a homogeneous set (having sufficient homogeneity) of URIs associated with close pages, to obtain a list of page URIs which are relevant relative to this set. It will be described later how to use this algorithm to obtain a set of pages relevant to an inhomogeneous set.
- Kernel a set K of reference URIs
- Authority a set A of candidate URIs
- the upstream homogeneity ws of a set S is defined as follows:
- ⁇ is an arbitrary constant lower but close to 1 (it serves to avoid divisions by zero but does not change the principle of the algorithm. If the set ⁇ is greater than K " then this constant can be equal to one
- l JP is equal to 1 if there is a link
- a p is the total of the pivot scores of the pages (of H) which point to at least one page of P (P being the current subset of S which is considered).
- this threshold could be a function of the current cardinality of H, in fact the closer we get to R '' (eg H ⁇ resort a ⁇ , the more the score calculated a chances of being correct already)
- the basic idea is to weaken the links between two pages that we know are close, by assigning a weight to each link, which weight will be equal to the complement of the proximity of the two linked pages (the greater the proximity, the more the link must be weakened). Once the links are thus weighted, it is possible to calculate the homogeneity of a set of pages using no longer the number of citing pages, but the sum of their weights.
- • / -x is the complement of the proximity between page P j and page P- if there is a link from page P j to page P, *, and zero otherwise • max (/ y , • x ⁇ ) is the complement of the proximity between the page P e H in question and the page P, e P for which the link between P and, presents the minimum proximity
- £ JP is equal to the complement of the proximity between page P, and page P, which is the closest to it and to which it has a link.
- ⁇ £ jP is the sum of the weights thus associated with the j pages of H which point to at least one of the pages of the subset P considered.
- the filtering described above uses a weight x. Since we now have the scores 51 of the citing pages, we can optionally improve the process by taking x ⁇ • h ⁇ as weight (instead of x), where h ⁇ is the score of the citing page (weakening a link (of a citing page P j to a quoted page P,) more when the score of the citing page P, is low It should be noted that to calculate the proximity x between two pages P, and P, linked, instead of using the equation of quantity of reasons as illustrated above, one can make the relation between:
- P N be the page of R, not yet in Kc, which has the highest relevance score. If its relevance score is lower than a fixed minimum score, return to point three (the current kernel is now complete). Otherwise insert it in Kc and go back to point four. Note that it will not be necessary to reset the pivot and authority scores, it is preferable to keep the last calculated values, so convergence should be very fast.
- the method for calculating relevance scores can be based on another equation
- the upstream and downstream processes can advantageously be integrated as follows: After the upstream processing (possibly even after each upstream iteration), to the candidate pages (R "+ ) having obtained a relevance score sufficient, we associate downstream a set of additional pages ("artificial pages") whose cardinality is a function of said relevance score. Each artificial page is also cited by (at least) one page of the request. the downstream an "advantage" to the scores of these good pages (of R '+ ) found by the upstream 54 , and consequently one also indirectly gives an advantage to the scores of the pages (of R +' + ) cited where appropriate by these good pages.
- the present process can be applied in addition to the existing methods in the prior art. Indeed, once the scores obtained for each page, one can artificially modify the respective numbers of the citing and cited pages before applying these methods.
- Figure 7 schematically presents such a process: the search for relevant pages can be applied recursively by extending the query with "Good pages found upstream”, “Good pages found downstream”, “Good pivot pages “And” Good authorities pages "which in the figure are framed by rectangles. With each recursion, the scores of the best pages found become a little lower (by the fact that the best pages found are added to the query each time) and the process stops when the scores cease to be sufficient.
- a system implementing the relative distillation method described above is capable of receiving a search request composed of a set of URIs allowing access to information resources such as web pages and providing in response the URIs (or directly pages) that are supposed to be the most relevant to that query.
- the request can for example consist of the user's favorite links, the aim of the system being for example to monitor the Web around these links and to notify the user when new interesting pages appear there, either in “Push” technology. at the initiative of a server, or in “Pull” technology at the initiative of the user.
- the user can of course directly provide the system with a set of URIs, however, other means can also be offered to assist him in preparing and submitting a search query.
- the user can use any of the following devices:
- a graphic object that can be activated for example by click eg a button
- URI hypertext links
- the system is provided with a means capable of switching the page to a state where each click on a link triggers the execution of a search query (containing this link).
- a key on the keyboard such as the "Ctrl" key, pressed while clicking (by a pointing device) is used to trigger the execution of a search query from the link on which cursor means pointing is positioned.
- the right mouse button (or equivalent) is used to trigger the execution of a search query from the link on which the mouse cursor is positioned.
- Each of these devices can advantageously make it possible to execute said search request in addition to (in parallel to) access to the page designated by the link in question.
- the result of the search query will for example be displayed in a second window (new browser instance) or in a browser pane 55 .
- this pane can be adjacent to the main pane in which the page containing the link that the user clicked was displayed and in which is then displayed the page accessed by clicking on this link.
- other URIs can be added automatically to search query 5 . These can in particular be:
- URIs that another user (“mentor” or refer) considers to be very relevant to the selected URI the mentor being determined automatically by the system, or specified by the user himself (chosen from a list of "friends” that he previously stored in the system), or even proposed by the page designer (the user can also choose from a list of "experts” proposed by the page designer).
- the user's action (as described above, for example clicking on a link with the right button and choosing the appropriate option) triggers the display of a accessory page in which: • in addition to the link that the user has just selected 58 , other links, which he has previously previously selected for this same request, are presented; o check boxes can be displayed in association with each link presented, so that the user can in particular select those that will actually form the request; • said accessory page is also provided with an input means (such as a button) making it possible to launch the search request.
- an input means such as a button
- Said accessory page may also contain unfoldable graphic objects (such as directories, lockers, files, or the like metaphor) representing requests in preparation other than the current request. The user can thus choose the request (s) which will be enriched by the new link he has just selected.
- one of the essential advantages of the system is that it can function (find the relevant information resources) even if the search query is made up of a plurality of URIs.
- said accessory page can be composed of two parts.
- One of these parts contains the elements described above (i.e. the elements of the request in preparation).
- the other part presents the content of the page designated by the link selected by the user.
- the server will returns said accessory page which thus includes:
- the fact of using the system represents an important advantage compared to conventional navigation on the Web: the user not only receives the page designated by the link he clicked (this is conventional navigation on the Web ), but at the same time it has the possibility of sending a request (containing several URIs) to obtain still other relevant resources in relation to this page.
- said accessory page is returned after a rapid (or even restricted) execution of the current search request to which the clicked link has been added.
- the second page then directly contains part of the result 62 .
- the user then receives not only the page designated by the link he clicked, but in addition he directly benefits from other relevant resources related to this page.
- said accessory page can be displayed in a pane adjacent to the main pane of the browser. This adjacent pane opens in response to the action of the user who wishes to display the request in preparation (ie said accessory page).
- the request in preparation can thus be displayed in parallel (asynchronously) with the display of the page designated by the clicked link; the latter appearing (independently) in the main pane.
- the result of the search query can then be presented in the same adjacent pane.
- a (partial) result can possibly be returned after partial or restricted execution of the current search request, request to which the clicked link has been added.
- the adjacent pane then directly presents a quick search result (which will eventually be completed later).
- the system can directly return the relevant URIs (or pages) already known and return the rest of the results offline.
- the server can advantageously already start to survey the Web (crawling in Anglo-Saxon terminology) —that is to say, constitute R ' , K, R ⁇ ⁇ , R, R and R + '+ as already described- around the selected link. b) Result of the execution of a search query
- the server can return the results directly (for example in return for the HTTP request) or deferred (for example by email).
- the server returns the URIs (resulting from a request) in a page having the same structure as said accessory page (or said request in preparation), namely:
- each URI 66 can thus be in at least one of the states following 67 : suggested (default state), accepted or deleted (URIs which are in the deleted state are not presented); • the page is provided with an input means (such as a button) making it possible to relaunch the search query.
- the returned page also presents the other requests (from the same user) in the form of expandable graphic objects, as already described.
- the presentation of these can be prioritized according to their relevance to the clicked link (according to the relevance calculation methods described below).
- the returned page presents control means allowing the user to create new requests and delete existing requests.
- the user can copy and paste URIs from existing requests or from any other resource.
- the user can move (ventilate) the URIs received in other requests.
- Each request can be accessed individually by means of its own URI.
- the presentation of the result of a search query includes the content of the pages (pointed to by the resulting URIs), for example in miniaturized form (thumbnails).
- each link can then be in one of the following states: suggested, accepted, deleted or frozen.
- R ⁇ * ⁇ , R + ⁇ , and i? + "+” are notably used to calculate the proximity of linked resources, and to filter, as described above, by taking the complement of this proximity as a weighting of the counting of the links in question. For each request (for example for each spot),
- the similarity of a resource to other resources is determined by comparing their contents. The following describes how to determine similarity based on the distribution of words in the resources in question.
- Time-dependent authority score Each new authority resource has a hypertext authority score (a ht ) and a similarity authority score (a s ). Let ⁇ be the relationship between
- ⁇ is therefore a number equal to 1 at the beginning of the life of a resource in the system, and decreases linearly until it reaches 0 when we say that the resource in question is old.
- ⁇ is used as a weighting to progressively move from a similarity score to a hypertext score and the formula for the overall score is
- a ⁇ a s + ⁇ 'a h ⁇
- (with ⁇ ' l- ⁇ ).
- the known process of absolute distillation on a set of nodes linked by links comprises the following steps: 1- to each node assign a pivot score equal to 1 as well as an authority score,
- step 4 4- iterate by repeating from step 2 until the algorithm converges, that is to say until the scores are no longer significantly different from the previous step.
- Steps 2 and 3 are replaced by the following:
- each node calculate its authority score by adding the pivotal scores of the nodes which point to it multiplied by the weight of the respective links, then normalize the authority scores so that their total is equal to 1,
- each node calculate its pivot score by adding the authority scores of the nodes to which it points multiplied by the weight of the respective links, then normalize the pivot scores so that their total is equal to 1,
- the weight of the similarity link between two resources is equal to the scalar product of their word distributions (i.e. to the sum, for each word found in the two resources, of the product of the frequencies of this word in these resources: the resulting sum is a number between zero - case where there is no word in common - and 1 - case where the two resources have the same content) after removing the non-significant words ("stop words" in Anglo-Saxon terminology).
- the methods described above also make it possible to select, from a set of additional resources, a resource which is the most relevant with respect to a starting resource.
- step (c) selection of additional resources (typically private resources again) most similar to the most relevant resources selected in step (b).
- Such a method notably makes it possible to dynamically generate the content of published web pages according to the context.
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
Description
Claims
Applications Claiming Priority (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
FR0200341 | 2002-01-11 | ||
FR0200341 | 2002-01-11 | ||
FR0205751 | 2002-05-07 | ||
FR0205751 | 2002-05-07 | ||
PCT/FR2003/000089 WO2003057648A2 (fr) | 2002-01-11 | 2003-01-13 | Procedes et systemes de recherche et d'association de ressources d'information telles que des pages web |
Publications (2)
Publication Number | Publication Date |
---|---|
EP1470501A2 true EP1470501A2 (fr) | 2004-10-27 |
EP1470501B1 EP1470501B1 (fr) | 2015-03-11 |
Family
ID=26213322
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
EP03709876.1A Expired - Lifetime EP1470501B1 (fr) | 2002-01-11 | 2003-01-13 | Procedes et systemes de recherche et d'association de ressources d'information telles que des pages web |
Country Status (4)
Country | Link |
---|---|
US (2) | US7676507B2 (fr) |
EP (1) | EP1470501B1 (fr) |
AU (1) | AU2003214311A1 (fr) |
WO (1) | WO2003057648A2 (fr) |
Families Citing this family (64)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7293228B1 (en) * | 1997-01-31 | 2007-11-06 | Timebase Pty Limited | Maltweb multi-axis viewing interface and higher level scoping |
AUPO489297A0 (en) * | 1997-01-31 | 1997-02-27 | Aunty Abha's Electronic Publishing Pty Ltd | A system for electronic publishing |
US20030041305A1 (en) * | 2001-07-18 | 2003-02-27 | Christoph Schnelle | Resilient data links |
US7363310B2 (en) | 2001-09-04 | 2008-04-22 | Timebase Pty Limited | Mapping of data from XML to SQL |
US7281206B2 (en) * | 2001-11-16 | 2007-10-09 | Timebase Pty Limited | Maintenance of a markup language document in a database |
US7281005B2 (en) * | 2003-10-20 | 2007-10-09 | Telenor Asa | Backward and forward non-normalized link weight analysis method, system, and computer program product |
US7343374B2 (en) * | 2004-03-29 | 2008-03-11 | Yahoo! Inc. | Computation of page authority weights using personalized bookmarks |
US7493320B2 (en) * | 2004-08-16 | 2009-02-17 | Telenor Asa | Method, system, and computer program product for ranking of documents using link analysis, with remedies for sinks |
US7577650B2 (en) * | 2005-04-13 | 2009-08-18 | Microsoft Corporation | Method and system for ranking objects of different object types |
US7496557B2 (en) * | 2005-09-30 | 2009-02-24 | International Business Machines Corporation | Microhubs and its applications |
US20070130120A1 (en) * | 2005-12-01 | 2007-06-07 | Yen-Fu Chen | System and method of selectively searching textual content |
US7921106B2 (en) * | 2006-08-03 | 2011-04-05 | Microsoft Corporation | Group-by attribute value in search results |
US20080071797A1 (en) * | 2006-09-15 | 2008-03-20 | Thornton Nathaniel L | System and method to calculate average link growth on search engines for a keyword |
US7912831B2 (en) * | 2006-10-03 | 2011-03-22 | Yahoo! Inc. | System and method for characterizing a web page using multiple anchor sets of web pages |
KR100705474B1 (ko) * | 2006-10-19 | 2007-04-09 | (주)이즈포유 | 웹 페이지에서 인디케이터를 이용한 정보 표시 및 관리시스템 |
TWI337712B (en) * | 2006-10-30 | 2011-02-21 | Inst Information Industry | Systems and methods for measuring behavior characteristics, and machine readable medium thereof |
US8661029B1 (en) | 2006-11-02 | 2014-02-25 | Google Inc. | Modifying search result ranking based on implicit user feedback |
US7809705B2 (en) * | 2007-02-13 | 2010-10-05 | Yahoo! Inc. | System and method for determining web page quality using collective inference based on local and global information |
US20080244428A1 (en) * | 2007-03-30 | 2008-10-02 | Yahoo! Inc. | Visually Emphasizing Query Results Based on Relevance Feedback |
US9092510B1 (en) | 2007-04-30 | 2015-07-28 | Google Inc. | Modifying search result ranking based on a temporal element of user feedback |
US20090055433A1 (en) * | 2007-07-25 | 2009-02-26 | Gerard Group International Llc | System, Apparatus and Method for Organizing Forecasting Event Data |
US20090132481A1 (en) * | 2007-11-15 | 2009-05-21 | Microsoft Corporation | Web site referral inference |
US7415460B1 (en) | 2007-12-10 | 2008-08-19 | International Business Machines Corporation | System and method to customize search engine results by picking documents |
US8515729B2 (en) * | 2008-03-31 | 2013-08-20 | Microsoft Corporation | User translated sites after provisioning |
US9183281B2 (en) * | 2008-04-18 | 2015-11-10 | International Business Machines Corporation | Context-based document unit recommendation for sensemaking tasks |
US20090271394A1 (en) * | 2008-04-29 | 2009-10-29 | Allen Thomas B | Determining the degree of relevance of entities and identities in an entity resolution system that maintains alert relevance |
US8015137B2 (en) | 2008-04-29 | 2011-09-06 | International Business Machines Corporation | Determining the degree of relevance of alerts in an entity resolution system over alert disposition lifecycle |
US8250637B2 (en) * | 2008-04-29 | 2012-08-21 | International Business Machines Corporation | Determining the degree of relevance of duplicate alerts in an entity resolution system |
US8326788B2 (en) * | 2008-04-29 | 2012-12-04 | International Business Machines Corporation | Determining the degree of relevance of alerts in an entity resolution system |
US9239884B2 (en) * | 2008-05-22 | 2016-01-19 | Alcatel Lucent | Electronic document processing with automatic generation of links to cited references |
US8515994B2 (en) * | 2008-06-12 | 2013-08-20 | International Business Machines Corporation | Reaching resource neighborhoods |
US20100070339A1 (en) * | 2008-09-15 | 2010-03-18 | Google Inc. | Associating an Entity with a Category |
US8512260B2 (en) * | 2008-10-29 | 2013-08-20 | The Regents Of The University Of Colorado, A Body Corporate | Statistical, noninvasive measurement of intracranial pressure |
US9569770B1 (en) | 2009-01-13 | 2017-02-14 | Amazon Technologies, Inc. | Generating constructed phrases |
US8423349B1 (en) * | 2009-01-13 | 2013-04-16 | Amazon Technologies, Inc. | Filtering phrases for an identifier |
US8768852B2 (en) * | 2009-01-13 | 2014-07-01 | Amazon Technologies, Inc. | Determining phrases related to other phrases |
US8706644B1 (en) | 2009-01-13 | 2014-04-22 | Amazon Technologies, Inc. | Mining phrases for association with a user |
US8706643B1 (en) | 2009-01-13 | 2014-04-22 | Amazon Technologies, Inc. | Generating and suggesting phrases |
US9009146B1 (en) | 2009-04-08 | 2015-04-14 | Google Inc. | Ranking search results based on similar queries |
US10108616B2 (en) * | 2009-07-17 | 2018-10-23 | International Business Machines Corporation | Probabilistic link strength reduction |
US8447760B1 (en) | 2009-07-20 | 2013-05-21 | Google Inc. | Generating a related set of documents for an initial set of documents |
US9298700B1 (en) | 2009-07-28 | 2016-03-29 | Amazon Technologies, Inc. | Determining similar phrases |
US10007712B1 (en) | 2009-08-20 | 2018-06-26 | Amazon Technologies, Inc. | Enforcing user-specified rules |
US9727842B2 (en) * | 2009-08-21 | 2017-08-08 | International Business Machines Corporation | Determining entity relevance by relationships to other relevant entities |
US8498974B1 (en) | 2009-08-31 | 2013-07-30 | Google Inc. | Refining search results |
US8972391B1 (en) | 2009-10-02 | 2015-03-03 | Google Inc. | Recent interest based relevance scoring |
US8799658B1 (en) | 2010-03-02 | 2014-08-05 | Amazon Technologies, Inc. | Sharing media items with pass phrases |
US8825860B2 (en) * | 2010-03-30 | 2014-09-02 | Qualcomm, Incorporated | Method and apparatus to facilitate voice activity detection and coexistence manager decisions |
US9129300B2 (en) * | 2010-04-21 | 2015-09-08 | Yahoo! Inc. | Using external sources for sponsored search AD selection |
WO2011156412A1 (fr) * | 2010-06-08 | 2011-12-15 | Optify, Inc. | Recommandation de possibilités de placement de lien |
US8332379B2 (en) * | 2010-06-11 | 2012-12-11 | International Business Machines Corporation | System and method for identifying content sensitive authorities from very large scale networks |
US9623119B1 (en) | 2010-06-29 | 2017-04-18 | Google Inc. | Accentuating search results |
US9002867B1 (en) | 2010-12-30 | 2015-04-07 | Google Inc. | Modifying ranking data based on document changes |
US8751488B2 (en) * | 2011-08-24 | 2014-06-10 | Waypart, Inc. | Part number search method and system |
US9009144B1 (en) * | 2012-02-23 | 2015-04-14 | Google Inc. | Dynamically identifying and removing potential stopwords from a local search query |
US9183312B2 (en) * | 2012-03-20 | 2015-11-10 | Google Inc. | Image display within web search results |
US8682892B1 (en) * | 2012-09-28 | 2014-03-25 | Google Inc. | Ranking search results |
US9558233B1 (en) * | 2012-11-30 | 2017-01-31 | Google Inc. | Determining a quality measure for a resource |
US11741090B1 (en) * | 2013-02-26 | 2023-08-29 | Richard Paiz | Site rank codex search patterns |
US9378299B1 (en) * | 2013-05-09 | 2016-06-28 | Google Inc. | Browsing pages in an electronic document |
US9363243B2 (en) * | 2014-03-26 | 2016-06-07 | Cisco Technology, Inc. | External indexing and search for a secure cloud collaboration system |
CN105528434A (zh) * | 2015-12-15 | 2016-04-27 | 无锡天脉聚源传媒科技有限公司 | 一种资源处理方法和装置 |
US10867321B1 (en) * | 2018-07-16 | 2020-12-15 | James D MacDonald-Korth | Automatic login link for targeted users without previous account creation |
US11188546B2 (en) * | 2019-09-24 | 2021-11-30 | International Business Machines Corporation | Pseudo real time communication system |
Family Cites Families (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5694594A (en) * | 1994-11-14 | 1997-12-02 | Chang; Daniel | System for linking hypermedia data objects in accordance with associations of source and destination data objects and similarity threshold without using keywords or link-difining terms |
US5870770A (en) * | 1995-06-07 | 1999-02-09 | Wolfe; Mark A. | Document research system and method for displaying citing documents |
US5920859A (en) * | 1997-02-05 | 1999-07-06 | Idd Enterprises, L.P. | Hypertext document retrieval system and method |
US6658623B1 (en) * | 1997-09-15 | 2003-12-02 | Fuji Xerox Co., Ltd. | Displaying in a first document a selectable link to a second document based on a passive query |
US6473752B1 (en) * | 1997-12-04 | 2002-10-29 | Micron Technology, Inc. | Method and system for locating documents based on previously accessed documents |
WO2001059625A1 (fr) * | 2000-02-10 | 2001-08-16 | Involve Technology, Llc | Systeme de creation et mise a jour d'une base de donnees d'informations sur les opinions des utilisateurs |
AU735010B3 (en) * | 2000-04-28 | 2001-06-28 | Forge Research Pty Limited | Business intelligence system |
US6463430B1 (en) * | 2000-07-10 | 2002-10-08 | Mohomine, Inc. | Devices and methods for generating and managing a database |
US6832218B1 (en) * | 2000-09-22 | 2004-12-14 | International Business Machines Corporation | System and method for associating search results |
US20020169851A1 (en) * | 2000-10-04 | 2002-11-14 | Robert Weathersby | Internet-based system for dynamically creating and delivering customized content within remote web pages |
JP2002222210A (ja) * | 2001-01-25 | 2002-08-09 | Hitachi Ltd | 文書検索システム、文書検索方法及び検索サーバ |
US8065620B2 (en) * | 2001-01-31 | 2011-11-22 | Computer Associates Think, Inc. | System and method for defining and presenting a composite web page |
US7386792B1 (en) * | 2001-03-07 | 2008-06-10 | Thomas Layne Bascom | System and method for collecting, storing, managing and providing categorized information related to a document object |
US7299270B2 (en) * | 2001-07-10 | 2007-11-20 | Lycos, Inc. | Inferring relations between internet objects that are not connected directly |
US6829606B2 (en) * | 2002-02-14 | 2004-12-07 | Infoglide Software Corporation | Similarity search engine for use with relational databases |
-
2003
- 2003-01-13 WO PCT/FR2003/000089 patent/WO2003057648A2/fr not_active Application Discontinuation
- 2003-01-13 EP EP03709876.1A patent/EP1470501B1/fr not_active Expired - Lifetime
- 2003-01-13 US US10/501,494 patent/US7676507B2/en not_active Expired - Fee Related
- 2003-01-13 AU AU2003214311A patent/AU2003214311A1/en not_active Abandoned
-
2010
- 2010-03-08 US US12/719,748 patent/US8290956B2/en not_active Expired - Lifetime
Non-Patent Citations (1)
Title |
---|
See references of WO03057648A3 * |
Also Published As
Publication number | Publication date |
---|---|
WO2003057648A2 (fr) | 2003-07-17 |
US7676507B2 (en) | 2010-03-09 |
US8290956B2 (en) | 2012-10-16 |
US20050171946A1 (en) | 2005-08-04 |
US20100228741A1 (en) | 2010-09-09 |
EP1470501B1 (fr) | 2015-03-11 |
AU2003214311A8 (en) | 2003-07-24 |
AU2003214311A1 (en) | 2003-07-24 |
WO2003057648A9 (fr) | 2004-07-29 |
WO2003057648A3 (fr) | 2004-06-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP1470501B1 (fr) | Procedes et systemes de recherche et d'association de ressources d'information telles que des pages web | |
US8589373B2 (en) | System and method for improved searching on the internet or similar networks and especially improved MetaNews and/or improved automatically generated newspapers | |
US7966316B2 (en) | Question type-sensitive answer summarization | |
US8918401B1 (en) | Systems and methods for providing searchable prior history | |
US20060155764A1 (en) | Personal online information management system | |
US20070255702A1 (en) | Search Engine | |
EP1719061A2 (fr) | Procédé mis en oeuvre dans un environnement informatique pour engendrer une vue courante à partir d'au moins un objet d'information source susceptible de varier | |
EP2174472A2 (fr) | Procede et dispositif de creation d'applications informatiques | |
WO2006026578A2 (fr) | Systeme est procede de gestion d'annotations en ligne | |
CA2727537A1 (fr) | Systeme et procede pour synthetiser des blogues | |
FR3043816B1 (fr) | Procede de suggestion de contenus extraits d’un ensemble de sources d’information | |
WO2002067142A2 (fr) | Dispositif d'extraction d'informations d'un texte a base de connaissances | |
Velásquez | Web site keywords: A methodology for improving gradually the web site text content | |
WO2001035269A2 (fr) | Systeme de partage d'informations entre au moins deux utilisateurs sur un reseau informatique | |
EP1334444A1 (fr) | Procede de recherche, de selection et de representation cartographique de pages web | |
WO2018015515A1 (fr) | Procedes de partage d'opinion, equipements et programmes d'ordinateur pour la mise en oeuvre des procedes | |
WO2001095146A2 (fr) | Systeme et procede permettant l'importation semi-automatique de fragments de ressources d'informations | |
WO2000068828A1 (fr) | Methode et systeme de creation de documents electroniques - auto-publiants et adaptatifs | |
Alaoui et al. | Semantic approach for the building of user profile for recommender system | |
WO2002029625A2 (fr) | Procede de surveillance de ressources d'information sur un reseau | |
FR3116355A1 (fr) | Détection d’au moins un thème partagé par une pluralité de documents textuels | |
FR2807182A1 (fr) | Systeme de traitement de donnees oriente objet a chargement progressif | |
EP3079076A1 (fr) | Procédé de détermination d'un gap sémantique, dispositif et programme correspondant | |
Liu et al. | Discovering and Exploiting Cumulative Cues for Informational Web Search | |
WO2001086496A1 (fr) | Systeme et procede de traitement de donnes pour l'acces cible a un serveur a partir d'une interrogation en langage naturel |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
17P | Request for examination filed |
Effective date: 20040823 |
|
AK | Designated contracting states |
Kind code of ref document: A2 Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LI LU MC NL PT SE SI SK TR |
|
AX | Request for extension of the european patent |
Extension state: AL LT LV MK RO |
|
17Q | First examination report despatched |
Effective date: 20080314 |
|
GRAP | Despatch of communication of intention to grant a patent |
Free format text: ORIGINAL CODE: EPIDOSNIGR1 |
|
INTG | Intention to grant announced |
Effective date: 20140724 |
|
GRAS | Grant fee paid |
Free format text: ORIGINAL CODE: EPIDOSNIGR3 |
|
GRAA | (expected) grant |
Free format text: ORIGINAL CODE: 0009210 |
|
AK | Designated contracting states |
Kind code of ref document: B1 Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LI LU MC NL PT SE SI SK TR |
|
REG | Reference to a national code |
Ref country code: GB Ref legal event code: FG4D Free format text: NOT ENGLISH |
|
REG | Reference to a national code |
Ref country code: CH Ref legal event code: EP |
|
REG | Reference to a national code |
Ref country code: IE Ref legal event code: FG4D Free format text: LANGUAGE OF EP DOCUMENT: FRENCH |
|
REG | Reference to a national code |
Ref country code: AT Ref legal event code: REF Ref document number: 715672 Country of ref document: AT Kind code of ref document: T Effective date: 20150415 |
|
REG | Reference to a national code |
Ref country code: DE Ref legal event code: R096 Ref document number: 60347394 Country of ref document: DE Effective date: 20150423 |
|
REG | Reference to a national code |
Ref country code: NL Ref legal event code: VDEP Effective date: 20150311 |
|
REG | Reference to a national code |
Ref country code: NL Ref legal event code: VDEP Effective date: 20150311 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: ES Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 Ref country code: FI Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 Ref country code: SE Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 |
|
REG | Reference to a national code |
Ref country code: AT Ref legal event code: MK05 Ref document number: 715672 Country of ref document: AT Kind code of ref document: T Effective date: 20150311 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: GR Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150612 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: NL Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: CZ Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 Ref country code: PT Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150713 Ref country code: SK Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 Ref country code: EE Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: AT Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 |
|
REG | Reference to a national code |
Ref country code: DE Ref legal event code: R097 Ref document number: 60347394 Country of ref document: DE |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: IT Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 |
|
PLBE | No opposition filed within time limit |
Free format text: ORIGINAL CODE: 0009261 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: DK Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 |
|
REG | Reference to a national code |
Ref country code: FR Ref legal event code: PLFP Year of fee payment: 14 |
|
26N | No opposition filed |
Effective date: 20151214 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: SI Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: BE Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20160131 |
|
PGFP | Annual fee paid to national office [announced via postgrant information from national office to epo] |
Ref country code: FR Payment date: 20160129 Year of fee payment: 14 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: LU Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20160113 |
|
REG | Reference to a national code |
Ref country code: CH Ref legal event code: PL |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: MC Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: CH Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20160131 Ref country code: LI Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20160131 |
|
PGFP | Annual fee paid to national office [announced via postgrant information from national office to epo] |
Ref country code: DE Payment date: 20160801 Year of fee payment: 14 Ref country code: GB Payment date: 20160729 Year of fee payment: 14 |
|
REG | Reference to a national code |
Ref country code: IE Ref legal event code: MM4A |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: IE Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20160113 |
|
REG | Reference to a national code |
Ref country code: DE Ref legal event code: R119 Ref document number: 60347394 Country of ref document: DE |
|
GBPC | Gb: european patent ceased through non-payment of renewal fee |
Effective date: 20170113 |
|
REG | Reference to a national code |
Ref country code: FR Ref legal event code: ST Effective date: 20170929 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: FR Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20170131 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: GB Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20170113 Ref country code: DE Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES Effective date: 20170801 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: CY Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 Ref country code: HU Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT; INVALID AB INITIO Effective date: 20030113 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: TR Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 |
|
PG25 | Lapsed in a contracting state [announced via postgrant information from national office to epo] |
Ref country code: BG Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT Effective date: 20150311 |